all_results.json

{
    "epoch": 3.0,
    "eval_logits/chosen": -4.902731895446777,
    "eval_logits/rejected": -4.834568977355957,
    "eval_logps/chosen": -402.26611328125,
    "eval_logps/rejected": -267.3836975097656,
    "eval_loss": 0.1433890014886856,
    "eval_rewards/accuracies": 0.916201114654541,
    "eval_rewards/chosen": 4.818836688995361,
    "eval_rewards/margins": 5.867211818695068,
    "eval_rewards/rejected": -1.0483758449554443,
    "eval_runtime": 328.4143,
    "eval_samples": 2862,
    "eval_samples_per_second": 8.715,
    "eval_steps_per_second": 0.545,
    "train_loss": 0.15624731566807995,
    "train_runtime": 68304.4343,
    "train_samples": 140201,
    "train_samples_per_second": 6.158,
    "train_steps_per_second": 0.096
}